Re: [CP2K:4615] help me I spent a lot of time to do these four things

[Marcella Iannuzzi]

Hi …

I would suggest that you proceed step by step first testing your set-up with simple energy and forces calculations.

For a simple bulk structure you should be able to converge the scf in a reasonable number of steps, and find that the forces on atoms are close to zero.

You should test the parameters in your input on efficiency and accuracy.

cp2k is a code for supercell calculations (no k-point). This means that it is thought for large simulation cells, including several replica of the unit cell.

Then please use a proper number of replica of the unit cell in the definition of the subsystem, and check the convergence with cell size, as you would do with a k-point calculation with the number of k-points in the mesh.

If you use scaled coordinates, add the SCALED keyword.

Since k-points are not implemented, the band structure cannot be calculated, but you can easily get the projected density of states.

Once you have well converged energy calculations, you can check the stress tensor to determine how far from equilibrium your system is.

The cell optimizer can be used in combination with the geometry optimizer if needed. 

As alternative, you can rescale box and coordinates and keep on performing total energy calculations to determine the energy-vs-volume curves

regards

Marcella

--------------------------------------------------------------
Marcella Iannuzzi              Phone : ++41 44 635 4479
Physical Chemistry Institute   FAX   : ++41 44 635 6838
University of Zurich               E-mail:  marcella@pci.uzh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
---------------------------------------------------------------

On Sep 18, 2013, at 3:28 AM, mohamed khuili wrote:

hi
all user CP2K, I spent a lot of time to install and use this code, I want to do four things:
- Optimize the cell
- The geometry optmiser
- Calculate the total energy of the system
- Have the density of states and band structure
you find below my input file, help me to achieve these four points, knowing that I utlise odrinateur my core I5

thank you very much

&FORCE_EVAL
 STRESS_TENSOR ANALYTICAL
 METHOD Quickstep
 &DFT
   BASIS_SET_FILE_NAME BASIS_MOLOPT
   POTENTIAL_FILE_NAME GTH_POTENTIALS
   &MGRID
     CUTOFF 90
   &END MGRID
   &QS
     METHOD GPW
   &END QS
   &SCF
     EPS_SCF 5.E-4
     MAX_SCF 50
     SCF_GUESS ATOMIC
     &OT
       MINIMIZER CG
     &END
   &END SCF
   &XC
     &XC_FUNCTIONAL PBE
     &END XC_FUNCTIONAL
   &END XC
 &END DFT
 &SUBSYS
   &CELL
        PERIODIC XYZ
        ABC 3.250 3.250 5.207
        ALPHA_BETA_GAMMA  90.0 90.0 120.0
        MULTIPLE_UNIT_CELL 2 2 2
   &END CELL
    &TOPOLOGY
       MULTIPLE_UNIT_CELL 2 2 2
       CONNECTIVITY OFF
    &END TOPOLOGY
 &COORD
   Zn     0.666670     0.333330     0.500000
   Zn     0.333333     0.666667     0.000000
   O      0.333333     0.666667     0.382500
   O      0.666667     0.333333     0.882500
   &END COORD

    &PRINT
      &ATOMIC_COORDINATES
      &END
      &CELL
      &END
    &END

    &KIND Zn
    BASIS_SET DZVP-MOLOPT-SR-GTH
    POTENTIAL GTH-PBE-q12
   &END KIND
   &KIND O
    BASIS_SET DZVP-MOLOPT-SR-GTH
    POTENTIAL GTH-PBE-q6
   &END KIND
 &END SUBSYS
&END FORCE_EVAL
 &MOTION
 &CELL_OPT
    KEEP_ANGLES .TRUE.
    MAX_DR 0.02
    RMS_DR 0.01
    MAX_FORCE 0.002
    RMS_FORCE 0.001
   OPTIMIZER CG
   &CG
    &LINE_SEARCH
     TYPE 2PNT
    &END LINE_SEARCH
   &END CG
 &END CELL_OPT
 &GEO_OPT
   OPTIMIZER BFGS
 &END GEO_OPT
 &END MOTION
&GLOBAL
 PROJECT ZnO
 RUN_TYPE CELL_OPT
 PRINT_LEVEL LOW
 EXTENDED_FFT_LENGTHS
&END GLOBAL

                          

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To post to this group, send email to cp...@googlegroups.com.
Visit this group at http://groups.google.com/group/cp2k.
For more options, visit https://groups.google.com/groups/opt_out.